[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.

22]

194 Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201

IF: 0.925
Asian Pacific Journal of Tropical Medicine
journal homepage: www.apjtm.org

doi: 10.4103/1995-7645.228433 ©2018 by the Asian Pacific Journal of Tropical Medicine. All rights reserved.

Mayaro virus infection, the next epidemic wave after Zika? Evolutionary
and structural analysis
Eleonora Cella1,2#, Marta Giovanetti3,4#, Teresa Milano5, Marta Fogolari2, Francesco Garilli6, Ivailo
Alexiev7, Riccardo Bazzardi8, Marco Salemi9, Luiz Carlos Junior Alcantara3, Silvia Angeletti2,
Stefano Pascarella5, Massimo Ciccozzi2
1
Public Health and Infectious Diseases, Sapienza University, Rome, Italy
2
Unit of Clinical Laboratory Science, University Campus Bio-Medico of Rome, Italy
3
Fundação Oswaldo Cruz, Salvador, Bahia, Brazil
4
University of Rome “Tor Vergata”, Rome, Italy
5
Dipartimento di Scienze Biochimiche "A. Rossi Fanelli", Università La Sapienza, 00185 Roma, Italy
6
Department of Medicine, University Campus Bio-Medico of Rome, Rome, Italy
7
National Reference Laboratory of HIV, National Center of Infectious and Parasitic Diseases, Sofia, Bulgaria
8
Department of Food Hygiene, Experimental Zooprophylactic Institute of Sardinia, Sassari, Italy
9
Emerging Pathogens Institute, University of Florida, Gainesville, FL, USA

A RT I C L E I N F O A B S T R AC T

Article history: Objective: To evaluate the evolution of the pathogen Mayaro virus, causing Mayaro fever (a
Received 15 November 2017 mosquito-borne disease) and to perform selective pressure analysis and homology modelling.
Received in revised form 28 December 2017 Methods: Nine different datasets were built, one for each protein (from protein C to non-structural
Accepted 1 February 2018 protein 4) and the last one for the complete genome. Selective pressure and homology modelling
Available online 2 March 2018
analyses were applied. Results: Two main clades (A and B) were pointed in the maximum
likelihood tree. The clade A included five Brazilian sequences sampled from 1955 to 2015. The
Keywords:
Brazilian sequence sampled in 2014 significantly clustered with the Haitian sequence sampled
Mayaro virus
in 2015. The clade B included the remaining 27 sequences sampled in the Central and Southern
Proteins
Evolutionary analysis America from 1957 to 2013. Selective pressure analysis revealed several sites under episodic
diversifying selection in envelope surface glycoprotein E1, non-structural protein 1 and non-
structural protein 3 with a posterior probability P曑0.01. Homology modelling showed different
sites modified by selective pressure and some protein-protein interaction sites at high interaction
propensity. Conclusion: Maximum likelihood analysis confirmed the Mayaro virus previous
circulation in Haiti and the successful spread to the Caribbean and USA. Selective pressure
analysis revealed a strong presence of negatively selected sites, suggesting a probable purging of
deleterious polymorphisms in functional genes. Homology model showed the position 31, under
selective pressure, located in the edge of the ADP-ribose binding site predicting to possess a high
potential of protein-protein interaction and suggesting the possible chance for a protective vaccine,
thus preventing Mayaro virus urbanization as with Chikungunya virus.

1. Introduction fever. The virus isolated in Trinidad in 1954 is a mosquito-borne

alphavirus reported in the South and Central America[1].
Mayaro virus (MAYV), belongs to Togaviridae family, triggers The genome is a single-strand RNA of 11.5 kb[2,3]. It has two
a febrile arthralgia syndrome close to Dengue and Chikungunya different open reading frames. The first one translates four
This is an open access article distributed under the terms of the Creative Commons
Attribution-NonCommercial-Share Alike 3.0 License, which allows others to remix, tweak
# and build upon the work non-commercially, as long as the author is credited and the new
These authors contributed equally to this work.
creations are licensed under the identical terms.
First author: Eleonora Cella, Public Health and Infectious Diseases, Sapienza University,
Rome, Italy. For reprints contact: reprints@medknow.com
E-mail: Eleonora.cella@yahoo.it ©2018 Asian Pacific Journal of Tropical Medicine Produced by Wolters Kluwer- Medknow

Corresponding author: Prof. Dr. Silvia Angeletti, Unit of Clinical Laboratory Science,
How to cite this article: Eleonora Cella, Marta Giovanetti, Teresa Milano, Marta Fogolari,
University Campus Bio-Medico of Rome, Via Alvaro del Portillo, 200, 00125 Rome, Italy.
E-mail: s.angeletti@unicampus.it Francesco Garilli, Ivailo Alexiev, et al. Mayaro virus infection, the next epidemic wave after
Prof. Massimo Ciccozzi, Unit of Clinical Laboratory Science, University Campus Bio- Zika? Evolutionary and structural analysis. Asian Pac J Trop Med 2018; 11(3): 194-201.
Medico of Rome, Via Alvaro del Portillo, 200, 00125 Rome, Italy.
E-mail: ciccozzi@iss.it
[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.22]
Eleonora Cella et al./Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201
non- structural proteins (NSP1–4), whereas the second open reading
frame encodes a single polyprotein, subsequently being processed
to five several proteins: capsid protein (C), two envelope surface
glycoproteins (E1 and E2), and two small peptides (E3 and 6k)[4,5].
Virus natural cycle involves non-human primates and Haemagogus
spp. mosquitoes in tropical areas of South America[6].
MAYV causes Mayaro fever (MF), a neglected endemic syndrome
of tropical Americas. After 7–12 d following a mosquito bite as
incubation time, patients develop fever, rash, headache, and arthralgia
that persist for several weeks[6]. The most prominent symptoms,
represented by joint pain and edema, are typical of the acute phase
of the disease but can persist chronically for several months[7,8].
Even though the disease is not severe and none case of death has
been reported, MF cause significant morbidity especially in rural
population[9], in rare case also with hemorrhagic manifestations[10].
The disease can determine temporary incapacitation to work and
hospitalization in some cases[1].
Two different genotypes have been described during MAYV
epidemics, the D genotyped prevalently isolated in Brazil and the L
genotype with a more wide distribution[1,10]. Reproducing the path
of transmission followed by Chikungunya virus (CHIKV), Dengue
virus (DENV) and Zika virus (ZIKV) in the Western countries
and Indian regions, MAYV has the potentiality to become a global
pathogen, due to its transmission mediated by urban mosquitoes
such as Aedes aegypti[11]. Similarly to CHIKV and ZIKV infection,
MAYV epidemics can be undiagnosed during DENV outbreaks
and incorrectly misdiagnosed as DENV infection in about 1%
of cases[12]. Case definition is clinical definition plus laboratory
diagnosis of MAYV[1].
MAYV identification is difficult to be achieved due to the relative
short duration of the viremic phase[10]. The available treatment for
MF is based on non-steroidal anti-inflammatory medications and
chloroquine, and no vaccine or specific antiviral drugs are available
at the moment[13]. For this reason, MAYV infection should represent
an emerging public health threat, and improved surveillance and
preventive measures should be provided to mitigate the already
burden on health systems in Latin America so as to limit the spread
of the virus infection.
In this study, the genetic diversity of MAYV has been investigated
with the aim to estimate the phylogenetic relationships between MAYV
strains circulating worldwide. Since MAYV is considered an emergent
pathogen, evolutionary analyses have been performed to obtain a
comprehensible overview positive-selection analyses and homology
modelling has been also performed to evaluate the pathogen evolution
and the virus evolution consequences in the protein recognition by host
immune response. Homology modelling of MAYV proteins has been
applied to allow mapping of sites under pressure and to put forward
structural hypotheses about possible consequences of mutations on
virus protein recognition by host immune response.
195
2. Materials and methods
2.1. Sequence dataset
Nine different datasets were assembled, eight for each protein
[C, E1, E2, E3, non-structural protein (NSP) 1, NSP2, NSP3, and
NSP4] and one for the complete genome, including sequences dated
from the year 1955 to the year 2015. All the datasets contained
33 sequences for each protein and the complete genome. All
the sequences and complete genome were downloaded by the
availability from the NCBI database (http://www.ncbi.nlm.nih.gov/)
and all the different genes were analyzed separately. All sequences
were aligned using MAFFT software v.7 and manually edited using
Bioedit software[14,15]. Modeltest software version 3.7[16] was used
to choose the best evolutionary model.
2.2. Phylogenetic signal and maximum likelihood analysis
The phylogenetic signal has been evaluated analyzing randomly
10 000 groups of four sequences (quartets), by the likelihood
mapping method using TREE-PUZZLE[17]. The dots inside the
equilater triangles correspond to the likelihood of the possible trees.
In the triangle, the corners represent the trees with fully resolved
topologies, the epicenter area represents the star-like phylogeny,
and the areas on the sides correspond to the network-like phylogeny
(presence of recombination or conflicting phylogenetic signals).
For substitution/saturation analysis assessment, the Xia’s test was
used with the transitions/transversions ratio vs. divergence graph in
DAMBE [http://dambe.bio.uottawa.ca/DAMBE/]. The percentage
of constant sites and parsimony-info sites were estimated using
MEGA7.
The maximum likelihood tree was inferred with the evolutionary
model previously selected with ModelTest, by using Phyml[18].
The phylogenetic tree was confirmed with the bootstrap analysis
(bootstrap values >75%) as reliability test for the branches.
2.3. Selective pressure analysis
Adaptive Evolution Server (http://www.datamonkey.org/) was
used to predict the ratio (氊) of non-synonymous to synonymous
substitution rates for each codon, with 氊>0 indicating diversifying
or positive selection[19]. The following selection analyses were
conducted according to the best-fit substitution model, also
determined within the Server: adaptive branch-site random effects
likelihood [aBSREL][20] model for the detection of lineage-specific
selection, fast unconstrained Bayesian approximation [FUBAR][21]
for inferring site-specific pervasive selection, Bayesian unrestricted
test for episodic diversifying selection [BUSTED][21] across the
region of interest, and the mixed effects model of evolution [MEME]
to identify episodic selection at individual sites[21]. Sites were
[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.22]

196 Eleonora Cella et al./Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201

considered to have been subjected to statistically significant positive NSP2 gene, 7.0% for NPS3 gene, 9.4% for NSP4 gene, and 1.2% for
or negative selection based on the following cut-offs: likelihood- the main dataset, complete genome region. All the datasets showed
ratio test (LRT) P曑0.05 for BUSTED and aBSREL and posterior sufficient phylogenetic signal (30%). All the genetic signal values and
probability (PP)] >0.90 for FUBAR, and LRT曑0.05 [with P曑0.01] phylogenetic information of the MAYV datasets are shown in Table
2. The percentage of the Parsimony-Info sites ranged from 1.74% (E2
for MEME. The reference sequence used to trace the amino acid
dataset)–19.65% (E1 dataset); instead the percentage of the constant
position found under selection was Accession Number: KX496990.1.
sites ranged from 73.62% (E1 dataset)–81.72% (NSP1 dataset).
In the maximum likelihood tree, two main clades (A and B) were
2.4. Homology modelling and structural analysis highlighted (Figure 2). The clade A included six sequences, sampled
from 1955 to 2015, of which five from Brazil and one from Haiti.
Homology modeling relied on the software Modeller version The Brazilian sequence sampled in 2014 clusters with the Haitian
9.17[22]. Template identification was carried out with the programs sequence sampled in 2015 with high bootstrap value (100%),
Blast [23] , HHPred [24] or Phyre2 [25] . Suitable templates were suggesting a very close phylogenetic relationship. The clade B
considered those structures sharing the highest sequence similarity included the remaining 27 sequences sampled in the Central and
to the target protein, solved at the best resolution available, and Southern America (Bolivia, Brazil, French Guiana, Peru, Trinidad
with maximum sequence coverage. Sequence alignments were and Tobago and Venezuela) from 1957 to 2013. Inside the clade
B, there is a statistically supported clade (B1) where one Peruvian
calculated with Clustal Omega[26] and were displayed or edited
sequence sampled in 2010 can be considered outgroup of the clade
with the Jalview editor[27]. Ten models of each target protein were
B1. Inside the clade B1, there are several statistically supported
built at the highest refinement level and that one showing the lowest
clusters, grouping mostly according the geographic location.
value of the Modeller target function, which is indicative of model
quality, was considered the representative model. ProsaII[28] and
3.2. Selective pressure analysis
Procheck[29] were utilized for model validation, while protein
structure visualization and analysis were carried out with the graphic Genetic variability was determined by nucleotide sequencing of
software PyMOL[30] or Chimera[31]. B-cell epitopes were predicted fragments ranging from 198 nt for the E3 protein to 2 394 nt for the
with a set of publicly available servers[32] listed in Table 1. Protein- NSP2 protein. The 毩 parameter of the 毭 distribution for all the
protein interaction site prediction used the meta-PPISP server. proteins analyzed was <1, demonstrating L-shape characteristic of
the distribution and proposing that rate heterogeneity of nucleotide
Table 1 substitution is across sites. Selective pressure analyses performed
B-cell epitope prediction for the Mayaro E1 protein variants. considering all the analyzed genomic regions (C, E1, E2, E3, NSP1,
Methods
Residue at position 300a)
Input typeb) NSP2, and NSP4) highlighted eight sites (107, 215, 243, 11, 231,
Leu Thr 497, 251, and 64) under strong negative selection (by using FUBAR)
ABCPred[33] + + sequence
indicated by 氊<0, suggesting high degree conservation in the
BCPred[34] - + sequence
Bepipred[35] - - sequence genomic region analyzed (Table 2). Using BUSTED, it has been
SVMtriP[36] - - sequence highlighted episodic diversifying selection only in E3 and NSP4.
Epitopia[37] - + sequence No branch-specific episodic diversifying selection was found using
BePRO[38] - - PDB aBSREL. Few sites with episodic diversifying selection were found
DiscoTope 2.0[39] - - PDB in E1, NSP1 and NSP3 with LRT P曑 0.01. There was one on each
EPCES[40] + - PDB
gene: amino acidic position 300 in E1, amino acidic position 242 in
EPSVR[41] +/- +/- PDB
ElliPro[42] + + PDB
NSP1 and amino acidic position 31 in NSP3 (using MEME).
CBTope[43] + + sequence
a)
+ and – denote predicted presence or absence of an epitope, respectively. b) 3.3. Homology modeling and structure analysis
PDB indicates homology model coordinates.
Homology model of the MAYV E1 protein (GenBank code
ANY58848) was built using the templates represented by the E1
3. Results envelope proteins from Semliki Forest virus (SFV) (PDB code
2ALA, resolution 3.0 Å) and CHIKV (PDB code 3N44, 2.35 Å). The
sequence alignment on which the modeling was based is displayed
3.1. Phylogenetic analysis and likelihood mapping
in Figure 3. MAYV E1 shares 76% and 60% sequence identity
As a first step, the phylogenetic signal for all the genomic regions with SFV and CHIKV proteins, respectively. Figure 4 reports the
superposition between the best MAYV E1 model and the structures
of the MAYV (C, E1, E2, E3, NSP1, NSP2, NSP3, NSP4, and
of the template proteins, and it also shows the location of sequence
complete genome) by likelihood mapping (Figures 1A–1I) was tested.
position 300 which was found to be under selective pressure.
The percentage of likelihood points in the central area of the triangles
This site is located in the domain III of the E1 protein within a 毬
was 12.9% for the first dataset (Capsid gene), 7.5% for E1 gene,
-strand of the immunoglobulin-like fold, before a conserved Cys
3.4% for E2 gene, 19.9% for E3 gene, 3.8% for NSP1 gene, 3.1% for
involved in a disulfide bond. A multiple sequence alignment of
[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.22]

Eleonora Cella et al./Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201
a set of 58 homologous E1 proteins from various alphaviruses
shows that position 300 can hold Ser (with frequency 58%), Glu
(18%), Lys (12%), Thr (5%), Leu (3%) or Ala (1%). Interestingly,
MAYV E1 displays a two-residue deletion in correspondence of
positions 344–345 of the SFV and CHIKV E1 proteins (Figure
3). This deletion occurs within a 毬-strand of the 毬-barrel of the
immunoglobulin fold and may induce a partial local conformational
rearrangement of the MAYV E1 loop encompassed by the sequence
positions 346–349 of domain III. Epitope prediction (Table 2) of
protein E1 variants which possess Thr or Leu at position 300 was
carried out using 11 methods listed in Table 1[32]. Epitope prediction
197
is quite inaccurate and for that reason a consensus approach is
generally recommended[32]. Overall, although no clear consensus is
evident, position 300 is predicted to be part of a region possessing
potential epitope propensity. In fact, the 5th and 6th methods out of
the 11 prediction methods employed indicate the occurrence of a
potential epitope with Leu or Thr at position 300, respectively (Table
2). Since the region bearing Thr is predicted as a potential epitope by
six methods, it may be conceived that the occurrence of this residue can
attribute a marginally higher epitope propensity. No suitable template
was found for the protein NSP1 and for that reason no homology
model could be built in our study. Modelling of the macrodomain

33.0% 29.5% 31.4%

33.6% 33.4% 33.0%

1.9

0.9
0.1

%
%

%
1.6

%
1.0
%
0.2

1.2% 7.5% 3.4%

32.5% 33.9% 32.5% 34.0% 34.0% 33.0%
31.9% 0.2% 33.4% 28.5% 2.1% 28.9% 32.2% 0.9% 30.2%

26.3% 26.5% 30.5%

33.5% 33.0% 32.8%

2.7

1.1
0.9
%

%
%
0.9

%
%

1.0
2.6
%

19.9% 12.9% 3.8%

32.8% 33.7% 32.5% 34.4% 34.1% 33.1%

25.6% 0.9% 25.5% 26.2% 2.9% 26.5% 31.7% 1.0% 31.0%

32.2% 29.7% 28.3%

33.7% 32.5% 32.7%

3.1%
0.6

1.0

2.0
%

%
%

%
0.7

1.0

1.6

7.0% 9.4%
32.7% 33.6% 33.2% 34.3% 32.8% 34.5%
31.1% 0.6% 31.6% 30.3% 1.0% 30.1% 28.3% 1.8% 28.6%

Figure 1. Likelihood mapping of all datasets.

Each dot represents the likelihoods of the three possible unrooted trees for a set of four sequences (quartets) selected randomly from the dataset: dots close to
the corners or the sides represent, respectively, tree-like, or network-like phylogenetic signal in the data. The central area of the likelihood map represents star-
like signal. The percentage of dots in the central area is given at the basis of each map.

Table 2
Evidence of selection pressure among codon sites in all the coding regions analyzed capsid, E1, E2, E3, NSP1, NSP2, NSP3 and NSP4 genes.
Analysis item Method Capsid E1 E2 E3 NSP1 NSP2 NSP3 NSP4
Number of sites with evidence of purifying selection FUBARa 107 215 243 11 231 497 251 64
Evidence of episodic diversifying selection BUSTEDb No No No Yes No No No Yes
MEMEc No 1 site: 300 No No 1 site: 242 No 1 site: 31 No
Branches aBSRELd 0 out of 47 0 out of 47 0 out of 49 0 out of 39 0 out of 63 0 out of 51 0 out of 43 0 out of 39
Fast unconstrained Bayesian approximation of pervasive selection (PP>0.90); bBayesian unconstrained test for episodic diversifying selection (LRT P曑0.05);
a

Site-specific episodic diversifying selection (LRT P曑0.01); dBranch-specific episodic diversifying (adaptive) selection (LRT P曑0.05).
c
[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.22]

198 Eleonora Cella et al./Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201

of MAYV NSP3 protein (Genbank code ALJ56198.1) utilized the

following templates: a fragment of the NSP from Sindbis virus (PDB
code 4GUA, resolution 2.85 Å); macrodomain from CHIKV (3GPG,
1.65 Å) and macrodomain of venezuelan equine encephalitis virus in
complex with ADP-ribose (3GQO, 2.60 Å). The complex between
MAYV NSP3 and ADP-ribose was predicted following the geometry
of the complex reported in 3GQO. MAYV NSP3 shares 52%, 66% and
56% sequence identity with 4GUA, 3GPG, and 3GQO, respectively.
According to the homology model, sequence position 31 of the MAYV
NSP3 is predicted at the mouth of the macrodomain binding site of the
Figure 4. Structural superposition among the MAYV E1 homology model
ADP-ribose (Figure 5) in proximity of residues (for example Asn24 and
(orange cartoon) and the homologous proteins from Semliki Forest virus
Val33) considered to be important for ADP-ribosylhydrolase activity
(grey) and Chikungunya virus (green).
in the homologous CHIKV NSP3 macrodomain[44]. Prediction of
Cyan subunit is the E2 protein from Chikungunya virus. Side chain of Leu at
protein-protein interaction sites attributed position 31 a high interaction
propensity (Figure 5). positions 300 of MAYV E1 is reported as red stick model. Disulfide bonds
are displayed as yellow sticks. Red areas in the 毬-strands of the Semliki
1961
1960 Forest and Chikungunya virus E1 proteins denote the two-residue insertions
2014
A 2015 found with respect to the MAYV E1 sequence.
1955
1991
1970
1995
2006
2004
2002
2006
2006
2006
2005
B1 2011
1995
2010
2010
2010
2010
2010
2010
B 2000
A B
2011
2011
1978
1978 Figure 5. Prediction of structural protein position.
1978
2013 [A] Structural superposition between the homology model of the MAYV
1957
2010
NSP3 macrodomain (light cyan cartoon 0 and the template macrodomain
Brazil Venezuela
Haiti from Venezuelan Equine Encephalitis virus [orange]. The ligand ADP-ribose
French Guiana
Peru is depicted by yellow sticks. Arrow indicates the position 31 occupied by an
Bolivia Trinidad and Tobago
Asp residue. [B] Surface of the MAYV macrodomain colored according to

Figure 2. Maximum likelihood tree of Mayaro virus. the protein-protein interaction potential using a scale ranging from blu (low

° along the branches indicating a statistical value from bootstrap (>70%). The potential) to red (high potential). Arrow marks the position 31. ADP-ribose is

legend for the location classification is in the left corner of the figure. displayed as in panel A.

10 20 30 40 50 60 70 80 90 100
HA177 105
2ALA 105
3N44\F 105

110 120 130 140 150 160 170 180 190 200
HA177 210
2ALA 210
3N44\F 210

220 230 240 250 260 270 280 290 300 310
HA177 315
2ALA 315
3N44\F 315

320 330 340 350 360 370 380

HA177 389
2ALA 384
3N44\F 391

Figure 3. Sequence alignment among the Mayaro E1 protein (labeled by HAITI) and the sequences of the E1 envelope proteins from Semliki Forest virus
(indicated by 2ALA) and Chykungunya virus (3N44 chain F).
Colors follow the Clustal scheme. Dashes mark sequence insertions/deletions.
[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.22]
Eleonora Cella et al./Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201
4. Discussion
MAYV has been reported in several tropical countries in the Central
and South America, as Brazil[1,45]. MAYV has the potentiality to
become a global pathogen, being its transmission mediated by urban
mosquitoes such as Aedes aegypti, as happened in the recent past for
CHIKV[46,47]. This analogy can also be supported by similar clinical
symptoms. Indeed the illness, which can cause chills, malaise,
headache, stomach pains and joint pains, is transmitted by that same
vector Aedes aegypti mosquito which transmits ZIKV, DENV, yellow
fever virus and CHIKV. As recently reported in CHIKV and ZIKV
epidemics, MAYV outbreaks can pass undiagnostigsate during
DENV epidemics[46,48]. Indeed the clinical symptoms of Mayaro are
so similar to Chikungunya and several other ailments that it is easy
to be misdiagnosed. Moreover dual infections from these viruses can
further complicate accurate diagnosis. MAYV is the one of the long
list of recently emerging viral pathogens that has received global
attention due to various outbreaks in tropical areas and its rapid
spread in several countries. Maximum likelihood analysis confirmed
the MAYV previous presence in Haiti and successful spread to areas
such as the Caribbean and USA[49].
Selective pressure analysis performed in this study revealed a
strong presence of negatively selected sites in all genomic regions
analyzed, suggesting a probable exclusion of deadly polymorphisms
in functionally genes. At the same time, this analysis highlighted a
few sites with episodic diversifying selection in E1, NSP1 and NSP3
with LRT P曑0.01 located in the amino acidic position 300 in E1,
amino acidic position 242 in NSP1 and amino acidic position 31 in
NSP3.
The contemporary lack of presence of high percentage of positively
selected sites also suggests and enforces the idea of highly adapted
phenotypes. Probably, the alternation cycle between arthropod vector
and human and no human hosts may impose a sort of barrier to non-
synonymous mutations in important genes. Amino acid conservancy
within MAYV sequences can be considered in combination with
potential immune escape amino acid mutations. The homology
modeling analysis showed localized high mutation frequency in the
structural and non-structural proteins (NS3 and NS5) corresponding
to a considerable alteration in the protein stability.
The position of the site 300 under positive selective pressure has
been mapped onto the predicted three-dimensional structure of
the MAYV E1 protein. In particular, model inspection indicates
that this site is located in the domain III of the E1 protein within
a 毬-strand of the immunoglobulin-like fold, before a conserved
Cys involved in a disulfide bond[50,51]. In the E1 envelope protein
of Sindbis virus, belonging to the Alphavirus genus, the position
equivalent to the MAYV E1 300 was proven to be part of a set of
transitional epitopes that become accessible only after virus particle
exposures to agents, such as heat, low pH and the like[50,51]. Change
199
in epitope accessibility was observed also upon virus particle
exposure to susceptible cells, which suggests that a conformational
modification of the envelope proteins takes place during cell binding.
Moreover, the structure site of the SFV E1 protein equivalent
to the MAYV Leu 300 is involved in E1–E1 contact within the
virus protein shell[52]. The high similarity between the MAYV E1
and the homologous protein sequences from other alphaviruses
suggests that the position 300 may have properties similar to
those attributed in the other viruses. Moreover, within the limits
of intrinsic inaccuracy, B-cell epitope prediction suggests that the
region surrounding the position 300 might have an immunogenic
potential. Interestingly, Thr at position 300 apparently may attribute
a higher epitope potential than the presence of Leu at the same
position. A multiple sequence alignment of a set of 58 homologous
E1 proteins from various alphaviruses shows that position 300 can
in other viruses hold other residues mostly polar such as Ser (with
frequency 58%). Furthermore, MAYV E1 displays a two-residue
deletion in correspondence of positions 344–345 of the SFV and
CHIKV E1 proteins. This deletion occurs within a 毬 -strand
of the 毬-barrel of the immunoglobulin fold and may induce a
partial local conformational rearrangement of the MAYV E1 loop
encompassed by the sequence positions 346–349 of domain III.
All these observations hint at the potential existence of an ongoing
fine process of MAYV E1 protein adaptation to cope with host
immunological surveillance.
MAYV protein NSP3 contains a N-terminal domain homologous to
the highly conserved macrodomain[53], which, in CHIKV, has been
demonstrated to possess ADP-ribosylhydrolase activity essential
for virus replication and virulence[44]. According to the homology
model, the position 31 found to be under selective pressure is located
in the edge of the ADP-ribose binding site and is predicted to possess
a high potential of protein-protein interaction. It may be conceived
that this position is involved in the recognition of the intracellular
targets of the MAYV macrodomain and consequently is subjected to
functional constraints. Currently, no licensed vaccines are available
for MF, and the only control strategy is based on human exposure to
potential vectors reduction. Two attempts of vaccine are described
in literature, one in human diploid cells and the other in murine
models[54]. A protective vaccine could decrease the potential risk of
MAYV urbanization through adaptative mutation that enhance the
vectorial ability of transmission, as happened with CHIKV[55] and
could be of great impact on public health.
MAYV has many undiscovered features of concern, and since it
was first isolated in Trinidad in 1954, it has been linked with limited
epidemics in Southern America[56] near the Amazon rainforest.
MAYV infection syndrome includes a group of symptoms such as
arthralgias, fever, headache, myalgias, rash, and sporadically nausea
and vomiting[57]. MAYV infections are underdiagnosed because of
a possible misleading with other mosquito-borne virus infections,
[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.22]
200 Eleonora Cella et al./Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201
especially dengue fever, endemic in common areas. Recently,
CHIKV has added disorientation in the diagnosis, especially due
to the prolonged arthralgia connected with CHIKV and MAYV
cases[56]. Our findings confirmed that MAYV is spreading in Brazil
since its isolation in the 1950s[49]. A serious concern arises from the
easy interaction and proclivity for MAYV/DENV co-infections. That
phenomenon should be the subject of careful studies and include
other arboviruses from the same geographic regions to clarify
the nature of the MAYV epidemic or other similar pathogens. In
addition, in light of these considerations, conducting molecular and
evolutionary studies are important to reveal the potential abilities
of this virus to trigger a significant epidemic, and help to elucidate
the MAYV way of transmission by Aedes and Haemagogus spp.
mosquitoes.
Conflict of interest statement
We declare that we have no conflict of interest.
References
[1] M
 ota MTdO, Ribeiro MR, Vedovello D, Nogueira ML. Mayaro virus: A
neglected arbovirus of the Americas. Fut Virol 2015; 10: 1109-1122.
[2] Gouandijka-Vasilache I, Manirakiza A, Gody JC, Banga-Mingo V,
Kongombe OO, Esona MD, et al. Rotavirus epidemiology in Bangui,
Central African Republic, 2008. Emerg Infect Dis 2014; 20(7): 1254-
1255.
[3] L
 avergne A, de Thoisy B, Lacoste V. Mayaro virus complete nucleotide
sequence and phylogenetic relationships with other alphaviruses. Virus
Res 2006; 117(2): 283-290.
[4] P owers A, Huang H, Roehring J, Strauss E, Weaver S. Family
Togaviridae. In: King AMQ, Lefkowitz E, Adams MJ, Carstens EB.
(editors) Virus taxonomy: Ninth Report of the International Committee on
Taxonomy of Viruses. San Diego: Elsevier Science; 2011, p. 1103-1110.
[5] Hallengard D, Kakoulidou M, Lulla A. Novel attenuated Chikungunya
vaccine candidates elicit protective immunityin C57BL/6 mice. J Virol
2014; 88(5): 2858-2866.
[6] Azevedo RS, Silva EV, Carvalho VL, Rodrigues SG, Nunes-Neto JP,
Monteiro H. Mayaro fever virus, Brazilian Amazon. Emerg Infect Dis
2009; 15(11): 1830-1832.
[7] Vasconcelos PF, Calisher CH. Emergence of human arboviral diseases in
the Americas, 2000–2016. Vector Borne Zoonot Dis 2016; 16(5): 295-301.
[8] Neumayr A, Gabriel M, Fritz J. Mayaro virus infection in traveller
returning from Amazon Basin, northern Peru. Emerg Infect Dis 2012;
18(4): 695-696.
[9] P
 inheiro FP, Freitas RB, Travassos da Rosa JF, Gabbay YB, Mello WA,
LeDuc JW. An outbreak of Mayaro virus diseasein Belterra, Brazil. I.
Clinical and virological findings. Am J Trop Med Hyg 1981; 30(3): 674-
681.
[10]Mourao MP, Bastos Mde S, de Figueiredo RP. Mayaro fever in the city of
Manaus Brazil, 2007–2008. Vector Borne Zoonot Dis 2012; 12(1): 42-46
[11]Long KC, Ziegler SA, Thangamani S. Experimental transmission of
Mayaro virus by Aedes aegypti. Am J Trop Med Hyg 2011; 85(4): 750-757.
[12]Forshey BM, Guevara C, Laguna-Torres VA, Cespedes M, Vargas J,
Gianella A. Arboviral etiologies of acute febrile illnesses in Western
South America, 2000–2007. PLoS Negl Trop Dis 2010; 4(8): e787.
[13]Taubitz W, Cramer JP, Kapaun A. Chikungunya fever intravelers: Clinical
presentation and course. Clin Infect Dis 2007; 45(1): e1-4.
[14]Katoh K, Kuma K, Toh H, Miyata T. MAFFT version 5: Improvement in
accuracy of multiple sequence alignment. Nucleic Acids Res 2005; 33(2):
511-518.
[15]Hall TA. BioEdit: A user-friendly biological sequence alignment editor
and analysis program for Windows 95/98/NT. Nucl Acids Symp 1999; 41:
95-98.
[16]Posada D, Crandall KA. MODELTEST: Testing the model of DNA
substitution. Bioinformatics 1998; 14(9): 817-818.
[17]Strimmer K, von Haeseler A. Likelihood-mapping: A simple method to
visualize phylogenetic content of a sequence alignment. Proc Natl Acad
Sci USA 1997; 94(13): 6815-6819.
[18]Guindon S, Dufayard JF, Lefort V, Anisimova M, Hordijk W, Gascuel
O. New algorithms and methods to estimate maximum-likelihood
phylogenies: Assessing the performance of PhyML 3.0. Syst Biol 2010;
59(3): 307-321.
[19]P ond SL, Frost SD, Muse SV. HyPhy:hypothesis testing using
phylogenies. Bioinformatics 2005; 21(5): 676-679.
[20]Kosakovsky Pond SL, Frost SD. Not so different after all: A comparison
of methods for detecting amino acid sites under selection. Mol Biol Evol
2005; 22(5): 1208-1222.
[21]Murrell B, Moola S, Mabona A, Weighill T, Sheward D, Kosakovsky
Pond SL, et al. FUBAR: A fast, unconstrained Bayesian approximation
for inferring selection. Mol Biol Evol 2013; 30(5): 1196-1205.
[22]Webb B, Sali A. Comparative protein structure modeling using
MODELLER. Curr Protoc Bioinformatics 2016; 54: 5.6.1-5.6.37.
[23]Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W,
et al. Gapped BLAST and PSI-BLAST: A new generation of protein
database search programs. Nucleic Acids Res 1997; 25(17): 3389-3402.
[24]Soding J, Biegert A, Lupas AN. The HHpred interactive server for protein
homology detection and structure prediction. Nucleic Acids Res 2005; 33:
W244-248.
[25]Kelley LA, Sternberg MJ. Protein structure prediction on the Web: A case
study using the Phyre server. Nature Protoc 2009; 4(3): 363-371.
[26]Sievers F, Higgins DG. Clustal omega. Curr Protoc Bioinformatics 2014;
48: 3.
[27]Waterhouse AM, Procter JB, Martin DM, Clamp M, Barton GJ. Jalview
Version 2--a multiple sequence alignment editor and analysis workbench.
Bioinformatics 2009; 25(9): 1189-1191.
[Downloaded free from http://www.apjtm.org on Saturday, April 13, 2019, IP: 10.232.74.22]
Eleonora Cella et al./Asian Pacific Journal of Tropical Medicine 2018; 11(3): 194-201
[28]Wiederstein M, Sippl MJ. ProSA-web: Interactive web service for the
recognition of errors in three-dimensional structures of proteins. Nucleic
Acids Res 2007; 35: W407-W410.
[29]Laskowski RA, Rullmannn JA, MacArthur MW, Kaptein R, Thornton
JM. AQUA and PROCHECK-NMR: Programs for checking the quality
of protein structures solved by NMR. J Biomol NMR 1996; 8(4): 477-
486.
[30]Schroedinger LLC. The PyMOL Molecular Graphics System 2015;
Version 1.8; 2015.
[31]Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM,
Meng EC, et al. UCSF Chimera--a visualization system for exploratory
research and analysis. J Comput Chem 2004; 25(13): 1605-1612.
[32]Potocnakova L, Bhide M, Pulzova LB. An introduction to B-cell epitope
mapping and in silico epitope prediction. J Immunol Res 2016; 2016:
6760830.
[33]Saha S, Raghava GP. Prediction of continuous B-cell epitopes in an
antigen using recurrent neural network. Proteins 2006; 65(1): 40-48.
[34]Chen J, Liu H, Yang J, Chou KC. Prediction of linear B-cell epitopes
using amino acid pair antigenicity scale. Amino Acids 2007; 33(3): 423-
428.
[35]Larsen JE, Lund O, Nielsen M. Improved method for predicting linear
B-cell epitopes. Immunome Res 2006; 2: 2
[36]Yao B, Zhang L, Liang S, Zhang C. SVMTriP: A method to predict
antigenic epitopes using support vector machine to integrate tri-peptide
similarity and propensity. PLoS One 2012; 7(9): e45152.
[37]Rubinstein ND, Mayrose I, Martz E, Pupko T. Epitopia: A web-server for
predicting B-cell epitopes. BMC Bioinformatics 2009; 10: 287.
[38]Sweredoski MJ, Baldi P. PEPITO: Improved discontinuous B-cell epitope
prediction using multiple distance thresholds and half sphere exposure.
Bioinformatics 2008; 24(12): 1459-1460.
[39]Kringelum JV, Lundegaard C, Lund O, Nielsen M. Reliable B cell
epitope predictions: Impacts of method development and improved
benchmarking. PLoS Comput Biol 2012; 8(12): e1002829.
[40]Liang S, Zheng D, Zhang C, Zacharias M. Prediction of antigenic
epitopes on protein surfaces by consensus scoring. BMC Bioinformatics
2009; 10: 302.
[41]Liang S, Zheng D, Standley DM, Yao B, Zacharias M, Zhang C. EPSVR
and EPMeta: Prediction of antigenic epitopes using support vector
regression and multiple server results. BMC Bioinformatics 2010; 11: 381.
[42]Ponomarenko J, Bui HH, Li W, Fusseder N, Bourne PE, Sette A, et
al. ElliPro: A new structure-based tool for the prediction of antibody
epitopes. BMC Bioinformatics 2008; 9: 514.
[43]Ansari HR, Raghava GP. Identification of conformational B-cell epitopes
in an antigen from its primary sequence. Immunome Res 2010; 6: 6
201
[44]McPherson RL, Abraham R, Sreekumar E, Ong SE, Cheng SJ, Baxter
VK, et al. ADP-ribosylhydrolase activity of Chikungunya virus
macrodomain is critical for virus replication and virulence. Proc Natl
Acad Sci U S A 2017; 114(7): 1666-1671.
[45]Vieira CJ, Silva DJ, Barreto ES. Detection of Mayaro virus infections
during a dengue outbreak in Mato Grosso,Brazil. Acta Trop 2015; 147:
12-16.
[46]Lo Presti A, Cella E, Angeletti S, Ciccozzi M. Molecular epidemiology,
evolution and phylogeny of Chikungunya virus: An updating review.
Infect Genet Evol 2016; 41: 270-278.
[47]Furuya-Kanamori L, Liang S, Milinovich G. Co-distribution and co-
infection of chikungunya and dengue viruses. BMC Infect Dis 2016; 16:
84.
[48]Giovanetti M, Milano T, Alcantara LC, Carcangiu L, Cella E, Lai A,
et al. Zika virus spreading in South America: Evolutionary analysis of
emerging neutralizing resistant Phe279Ser strains. Asian Pac J Trop Med
2016; 9(5): 445-452.
[49]Mavian C, Rife BD, Dollar JJ, Cella E, Ciccozzi M, Prosperi MCF, et al.
Emergence of recombinant Mayaro virus strains from the Amazon basin.
Sci Rep 2017 7(1): 8718.
[50]Voss JE, Vaney MC, Duquerroy S, Vonrhein C, Girard-Blanc C, Crublet
E, et al. Glycoprotein organization of Chikungunya virus particles
revealed by X-ray crystallography. Nature 2010; 468(7324): 709-712.
[51]M ukhopadhyay S, Zhang W, Gabler S, Chipman PR, Strauss EG,
Strauss JH, et al. Mapping the structure and function of the E1 and E2
glycoproteins in alphaviruses. Structure 2006; 14(1): 63-73.
[52]Meyer WJ, Johnston RE. Structural rearrangement of infecting Sindbis
virions at the cell surface: Mapping of newly accessible epitopes. J Virol
1993; 67(9): 5117-5125.
[53]Roussel A, Lescar J, Vaney MC, Wengler G, Wengler G, Rey FA.
Structure and interactions at the viral surface of the envelope protein E1
of Semliki Forest virus. Structure 2006; 14(1): 75-86.
[54]Rack JG, Perina D, Ahel I. Macrodomains: Structure, function, evolution,
and catalytic activities. Annu Rev Biochem 2016; 85: 431-454.
[55]Robinson DM, Cole FE Jr, McManus AT, Pedersen CE Jr. Inactivated
Mayaro vaccine produced in human diploid cell cultures. Mil Med 1976;
141(3): 163-166.
[56]Tsetsarkin KA, Chen R, Leal G, Forrester N, Higgs S, Huang J, et al.
Chikungunya virus emergence is constrained in Asia by lineage-specific
adaptive landscapes. Proc Natl Acad Sci U S A 2011; 108(19): 7872-7877.
[57]Tesh RB, Watts DM, Russell KL, Damodaran C, Calampa C, Cabezas
C, et al. Mayaro virus disease: An emerging mosquito-borne zoonosis in
tropical South America. Clin Infect Dis 1999; 28(1): 67-73.